[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <d70b62f8-e24e-4a10-8c6b-18d2e19f1337@collabora.com>
Date: Thu, 20 Mar 2025 13:37:49 +0100
From: AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>
To: Alexandre Mergnat <amergnat@...libre.com>,
Catalin Marinas <catalin.marinas@....com>, Will Deacon <will@...nel.org>,
Neil Armstrong <neil.armstrong@...aro.org>,
Jessica Zhang <quic_jesszhan@...cinc.com>,
Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
Maxime Ripard <mripard@...nel.org>, Thomas Zimmermann <tzimmermann@...e.de>,
David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
Chun-Kuang Hu <chunkuang.hu@...nel.org>,
Philipp Zabel <p.zabel@...gutronix.de>,
Matthias Brugger <matthias.bgg@...il.com>
Cc: linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
dri-devel@...ts.freedesktop.org, linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH v8 2/3] drm/panel: startek-kd070fhfid015: add another init
step
Il 20/03/25 09:48, Alexandre Mergnat ha scritto:
> Currently, the panel set power, set gpio and enable the display link
> in stk_panel_prepare, pointed by drm_panel_funcs.prepare, called by
> panel_bridge_atomic_pre_enable, pointed by
> drm_bridge_funcs.atomic_pre_enable. According to the drm_bridge.h,
> atomic_pre_enable must not enable the display link
>
> Since the DSI driver is properly inited by the DRM, the panel try to
> communicate with the panel before DSI is powered on.
>
The panel driver shall still be able to send commands in the .prepare() callback
and if this is not happening anymore... well, there's a problem!
> To solve that, use stk_panel_enable to enable the display link because
> it's called after the mtk_dsi_bridge_atomic_pre_enable which is power
> on the DSI.
>
> Signed-off-by: Alexandre Mergnat <amergnat@...libre.com>
> ---
> .../gpu/drm/panel/panel-startek-kd070fhfid015.c | 25 +++++++++++++---------
> 1 file changed, 15 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/gpu/drm/panel/panel-startek-kd070fhfid015.c b/drivers/gpu/drm/panel/panel-startek-kd070fhfid015.c
> index c0c95355b7435..bc3c4038bf4f5 100644
> --- a/drivers/gpu/drm/panel/panel-startek-kd070fhfid015.c
> +++ b/drivers/gpu/drm/panel/panel-startek-kd070fhfid015.c
> @@ -135,19 +135,9 @@ static int stk_panel_prepare(struct drm_panel *panel)
> gpiod_set_value(stk->enable_gpio, 1);
> mdelay(20);
> gpiod_set_value(stk->reset_gpio, 1);
> - mdelay(10);
> - ret = stk_panel_init(stk);
> - if (ret < 0)
> - goto poweroff;
Also, you're moving both init and set_display_on to the enable callback...
this is suboptimal.
You should do the DrIC setup in .prepare() (can include SLEEP OUT), and then you
should have a .enable() callback that calls DISP ON, a .disable() callback that
calls DISP OFF, and .unprepare() that turns everything off.
Cheers,
Angelo
> -
> - ret = stk_panel_on(stk);
> - if (ret < 0)
> - goto poweroff;
>
> return 0;
>
> -poweroff:
> - regulator_disable(stk->supplies[POWER].consumer);
> iovccoff:
> regulator_disable(stk->supplies[IOVCC].consumer);
> gpiod_set_value(stk->reset_gpio, 0);
> @@ -156,6 +146,20 @@ static int stk_panel_prepare(struct drm_panel *panel)
> return ret;
> }
>
> +static int stk_panel_enable(struct drm_panel *panel)
> +{
> + struct stk_panel *stk = to_stk_panel(panel);
> + int ret;
> +
> + ret = stk_panel_init(stk);
> + if (ret < 0)
> + return ret;
> +
> + ret = stk_panel_on(stk);
> +
> + return ret;
> +}
> +
> static const struct drm_display_mode default_mode = {
> .clock = 163204,
> .hdisplay = 1200,
> @@ -239,6 +243,7 @@ drm_panel_create_dsi_backlight(struct mipi_dsi_device *dsi)
> }
>
> static const struct drm_panel_funcs stk_panel_funcs = {
> + .enable = stk_panel_enable,
> .unprepare = stk_panel_unprepare,
> .prepare = stk_panel_prepare,
> .get_modes = stk_panel_get_modes,
>
Powered by blists - more mailing lists